Kraal, Pieter J.
Citation
Kraal, P. J. (2005, October 20). A grammar of Makonde (Chinnima, Tanzania). Retrieved from
https://hdl.handle.net/1887/4271
Version:
Corrected Publisher’s Version
License:
Licence agreement concerning inclusion of doctoral thesis in the
Institutional Repository of the University of Leiden
Downloaded from:
https://hdl.handle.net/1887/4271
1.1 The Makonde people: demography and geography
1.2 The Makonde language: classification and dialectal variation 1.3 Field work
1.4 About this grammar: background, previous studies and theory
7KH0DNRQGHSHRSOH GHPRJUDSK\DQGJHRJUDSK\
The Makonde live on both sides of the river Ruvuma, which forms the border between Tanzania and Mozambique. The (WKQRORJXH (Grimes 2000) gives the
number of Makonde speakers as 900,000 for Tanzania and 360,000 for Mozambique; these figures relate to 1993. The plateaus, on which most Makonde live, rise either side of the Ruvuma from sea level to an altitude of ca. 900m. The Makonde of Mozambique live in the north-east of the Province of Cabo Delgado, the majority of the Makonde of Tanzania live in two of the three districts of the Mtwara Region: Mtwara and Newala. In the third district, Masasi, Makonde live next to Makua and Yao. The northern neighbours of the Makonde in the Lindi Region are the Mwera. The Makonde plateau in Tanzania extends about 180 km from east to west, and about 80 km from south to north. It is a dry area, covered with dense bush and scattered high trees. There are a number of forest reserves and two lakes, and the highest mountains are situated in the north-west. 0DNRQGH means ‘dry areas’, and there is a
singular form /LNRQGH, which is also the name of several villages on the plateau.
The Makonde of Tanzania are divided into three main groups: the Nnima people who live in the north-west of the plateau, the Ndonde people who live on the south-west of the plateau, and the Maraba people who live near the Ruvuma and along the coast as well as on the eastern part of the plateau. The Maraba are part of the Swahili coastal culture, which next to their Makonde origin determines their cultural and historical identity, reflected in their religion (Islam), their fishing practices, the way they dress and their language, which is strongly influenced by Swahili. There is another group that claim to be part of the Makonde people: the Matambwe. They live near the Ruvuma next to the Ndonde with whom they intermarry.
Each Makonde group consists of several clans, which in turn consist of several extended families. The most important man of the family used to be the QMRPED, the
elder brother of the mother. Clans are headed by a chief who is appointed by his predecessor who normally chooses his sister’s son called PZLSZDZH. Amwenye
Mandanda, the father of our first informant, was chief of the Lukanga clan until he died in 1989; he was a MXPEH, a local leader appointed by the British colonial
Most Makonde are farmers who live in small villages. They grow mainly rice, maize and cassava. Those who live along the coast (the Maraba) and near the Ruvuma (the Maraba and the Matambwe) also practice fishery. Most of them are either Muslim or Christian (Catholic, Lutheran, Evangelical), many practice their traditional religion as well. In the past, Matambwe women used to wear decorative little round spots on both cheeks and the forehead as well as paintings on the chest; Makonde women used to wear a lip plug in their upper lip (like the Makonde of Mozambique, the Mwera and the Makua), Yao women used to wear a small pin in their right nostril, while the Makonde of Mozambique used to tattoo their faces as well as sharpened teeth. These practices are still found today, but much less frequently. The following so-called teasing relationships (Swahili XWDQL, Makonde XYLOR) exist in the area: Yao–Mwera,
Makua–Matumbi and Makonde–Ngoni.
The wood carvings for which the Makonde are known throughout the world are mainly produced by the Makonde of Mozambique. The black wood from which the carvings are made comes from a tree called PSLQJR in Swahili as well as in Makonde.
7KH0DNRQGHODQJXDJH FODVVLILFDWLRQDQGGLDOHFWDO
YDULDWLRQ
Makonde is a Bantu language, P.23 in Guthrie’ s referential classification (Guthrie 1948). In this classification, the name Makonde refers only to the Makonde of Tanzania. The other languages of Guthrie’ s “Yao Group” P.20 are Yao (P.21), Mwera (P.22), Ndonde (P.24) and Mabiha or Mavia (P.25).
As is clear from the descriptions, Mavia (P.25) is in fact the Makonde language of Mozambique (Shimakonde); it is an abusive term meaning ‘savage people’ and heavily disliked by the Makonde of Mozambique, but found widely in the literature. Our research revealed that Ndonde (P.24) does not exist as a separate language. The name Ndonde means ‘lower parts (of the plateau)’ or ‘western direction’ , and it is the name of one of the dialects of the Makonde of Tanzania as well as the name of one of the dialects of the Makonde of Mozambique; both dialects are spoken in the lower plateau areas in Tanzania and Mozambique. Ndonde is also the name of an area around Masasi to the west of the plateau in Tanzania. This is probably the reason why language maps of Tanzania give the name “Ndonde” to different locations: either south of the plateau (on the “Tervuren Map” as used in $IULFDQD/LQJXLVWLFDsince
1965) or to the west of it (Polomé and Hill 1980, p.30). Informants also reported that there are people called Ndonde living near Liwale (Lindi Region) as well as along the coast; neither speech variant forms part of the Makonde language but are rather part of the language of the Ngindo people.
Kimachinga. Harries (1940), talking about the linguistic situation in Tanzania, states that Matambwe, Ndonde and Makonde are three separate though related languages. Nurse (1979) mentions five Makonde dialects in his description of sample Bantu languages of Tanzania: Chinnima, Chimahuta, Kimaraba, Kimatambwe and Kimaviha. Odden (1990), in two articles describing two Makonde dialects in Tanzania, follows Nurse (p.c.) and mentions the same five dialects: Chimaraba, Chimatambwe, Chinnima, Chimahuta and Chimaviha. Our research of the Makonde language of Tanzania (called Chimakonde) revealed four dialects: Chimaraba, Chimatambwe, Chinnima, and Chindonde.
Our own research indicates that both Matambwe and Ndonde are dialects of Chimakonde. We differ from Lorenz in that he did not recognize Chinnima as a Makonde dialect, that he considers Chindonde to be a separate language (which is very common in the older literature), and that he considers Machinga to be a Makonde dialect. We differ from Nurse and Odden in that they applied the name Chimahuta to what we call Chindonde (according to all our informants, Chimahuta does not exist as a separate dialect), and that they mentioned Kimaviha or Chimaviha, which is in fact the Makonde language of Mozambique.
Liebenow (1971) reports that Makonde people, talking about their origin and past, mention Ndonde as the place west of the plateau around Masasi where a section of the Makonde people gathered in the first half of the nineteenth century, and where they stayed for about 20 years before moving onto the plateau in order to escape attacks from the Ngoni (who were also called Mafiti, Maviti or Mafita). Before this time, they lived together with the other Makonde groups around Negomano in northern Mozambique, where they arrived after having left what they consider to be their place of origin: (again) Ndonde, south-west of Negomano in Mozambique. Lichelo hill (with a summit called Mkundi) is sometimes given as their exact place of birth, while others mention Mkula. Dias (1964) reports that the original Makonde home south-west of Negomano lies in an area south of lake Malawi (Nyassa) according to the tradition.
came about 50 years later to the area according to Halimoja, who also mentions the Ndonde north of Masasi, probably the Ndonde who now live near Liwale (see above). Our Matambwe informant, Leonardi Petro Milanzi, reported that almost all Matambwe live in Tanzania, only some live in Mozambique; they all live near the Ruvuma, and they are fishermen. The word Matambwe refers to ‘mist’ or ‘haze’ , which often hangs over the Ruvuma banks. The Matambwe have some characteristics which set them apart as an ethnic group: they are recognized as a distinct group of people with a special profession (fishing, they do not cultivate), they have their own habits, practices and ways of decorating their bodies, they live in a special area, and they have their own language. Milanzi said that their language is very close to Chimakonde, with a high level of mutual intelligability, though their way of speech and some vocabulary items are different.
According to Milanzi, there are three distinct groups within the Matambwe people. One group lives south of Tunduru (Ruvuma Region) in the Yao area, and their language is a bit like the Yao language. A second group lives to the east of the first group in Masasi District and Newala District (Mtwara Region), and their language is very like Chimakonde. The third group lives in between. Milanzi said that this third group is also called Ndonde, their language is Matambwe but also called Ndonde, and it differs a bit from the language of the second group. Andrew Clayton, an anthropologist who worked on the Makonde plateau of Tanzania in the early nineties, suggests (p.c.) that the second and third group mentioned by Milanzi are one and the same group. He further says that the Makonde living on the southern part of the plateau are known as Ndonde people, and that most people regard the Ndonde people as the same as the Matambwe people. Others insist that these two groups are different, among them our Ndonde informant, Mariamu Mshamu. We think that they are different groups, although they live closely together and frequently intermarry. The Matambwe language has become so close to Makonde that it may be considered as one of its dialects.
there called Makwe whose language is closer to Chimaraba of Tanzania than to Shimakonde of Mozambique.
In Tanzania, in Mtwara District, the locations which are mentioned as Maraba centers are Msimbati, Msangamkuu, Naumbu, Mkungu and Mgao. Mgao is known in the literature because of Kimgao, the old Swahili dialect, spoken from Kilwa southwards, from the fourteenth century until it disappeared in the nineteenth century when it was replaced by the Kiunguja variant of Swahili. When the Maraba came to Tanzania (including Mgao) at the end of the eighteenth century, their language probably came under the influence of Kimgao. We had hoped to find traces of Kimgao, but during our short visits to the village of Mgao, we concluded that Kimgao has completely been replaced by Kiunguja, as has occurred elsewhere. The only trace may be in the name: people speak of Kimgao when talking about the Maraba as spoken in Mgao, while they do not refer to the Maraba of other places in the same way, e.g., the Maraba of Naumbu is not generally called Kinaumbu, but rather Kimaraba, Kimwamba or Kimwambao, three names used for the speech of the whole area. The women of Mgao, who habitually speak Maraba while the men habitually speak Swahili, call their village Kumwalu and their language Chimwalu. The form PZDOX is
the Maraba equivalent of Swahili PJDR.
The second migration route brought Makonde to the plateau in Tanzania near Nambunga. One of the places there is Mahuta, about which there is a legend which explains why the Makonde (formerly the Matanda people) live on a plateau where there is no water; the legend tells about a man from Mahuta who carves a piece of wood which becomes a woman. We conclude that this second migration brought the Ndonde people to the southern part of the plateau; they speak the Chindonde variant of Makonde. We also think that the Matambwe people took part in this migration since they were found south of Masasi at the beginning of the nineteenth century according to Halimoja (see above).
According to Liebenow’ s report, inhabitants of Mahuta and Namahonga say that their ancestors arrived earlier than those of the third migration route, who approached the plateau from the north near Mnima via Luchemo near Majembe. These are the Makonde who first went to the Ndonde area around Masasi where they stayed for about 20 years before moving onto the plateau. We conclude that this third migration brought the Nnima people to the north-eastern part of the plateau. As this part of the plateau lies at high altitude and has several hills, the people are called Nnima, which means ‘height’ . My first major informant, the late Mariamu Mandanda, described the Nnima area as a circle of villages, mainly located in Kitangari Division, Newala District: Kitangari, Nnyambe, Malatu, Majembe, Nanbunga, Chilangala, Minjale, Mkoma, Nkudumba, Nakahako, Luchemo and Mkalenda. Liebenow finally mentions a fourth migration route which is still going on: Makonde from the plateau move to the coast.
by P. H. Gulliver (Berry 1971:113), Machinga and Matambwe (called Matumbi, different from P.13 Matumbi which is also mentioned there) are called southern Tanzania Bantu languages. The same names appear in Sutton’ s list (1968:80). Neither Machinga nor Matambwe are mentioned in the earlier /LQJXLVWLF %LEOLRJUDSK\ by
Whiteley and Gutkind (1958), nor in Bryan’ s +DQGERRN of the Bantu languages
(1959).
Machinga means ‘mountains’ . The Machinga people live between Lindi (Lindi Region) and Kilwa Kivinje in a mountainous area inland from the coast. A place often mentioned in connection with Machinga is Mchinga, north of Lindi. Although the British colonial government considered the Machinga as a separate ethnic group, the Maraba who live there along the coast consider them to be Maraba who have become mixed with Mwera people, though their language is much like Maraba. Some Mwera people told us that the Machinga are Mwera and their language is closer to Mwera than to Maraba. Harries (1950), in his Mwera grammar, mentions three dialects, one of which is Chimachinga (or Chindandawala) spoken south of the Mbemkuru river. Polomé and Hill (1980) also note that some of their informants considered Machinga as part of Mwera. However, we do not have data to decide whether Chimachinga is a dialect of Makonde or Mwera, and we conclude that more research is needed. (There is also a dialect of the Yao language which is called Chimachinga, spoken in the western part of Masasi District. We got information about this dialect from Immanueli Gambe, born 1922 in the Masasi District, via Mathias Mwaya, born 1958 in the Mtwara District.)
language, or that Kimwani (as well as Chimaraba and Kimakwe) are dialects of Swahili or (Swahilized) dialects of Makonde. Our investigation is very clear in its conclusion with respect to Chimaraba and Kimakwe: the final option is the right one. Chimaraba and Kimakwe are Makonde dialects because most parts of their grammar are basically Makonde. We can not go into detail, but to mention some examples: the locative classes, most tenses, the conjoint/disjoint distinction, their prosodic structure and their tonal system are clearly Makonde. Exactly these arguments are decisive in solving the question whether Chimakonde (Tanzania) and Shimakonde (Mozambique) are both variants of the same language or can best be considered to be separate languages: as both make use of the same basic instruments as mentioned above in building their language, they can be considered to be variants of Makonde. Since we have not seen enough data of Kimwani to draw definite conclusions, we are less sure of its dialectal status. (According to Benjamin Leach of SIL (p.c.), the Mwani people themselves would disagree strongly that they form part of the Makonde people, or that their language is a dialect of Shimakonde.)
We now come with an overview of the variants that belong to the Makonde (P.23) group. As we have said earlier in this section, Ndonde (P.24) does not exist as a separate language, but it is the name of dialects of Chimakonde and Shimakonde; Mavia (P.25) is the same as Shimakonde.
0DNRQGH
&KLPDNRQGH 6KLPDNRQGH
Chinnima Shimakonde do Planalto Chindonde Shindonde Chimaraba Shiyanga Kimakwe Shimwalu Chimatambwe Shimwambe Chimachinga? Simakonde Kimwani?
)LHOGZRUN
Makonde of Mozambique, come from Mozambique, Mueda District. We start with the informants of Chinnima, the dialect of the Makonde of Tanzania to which the main body of this thesis is devoted.
Chinnima, main informants:
Ms Mariamu Mandanda, born 1942 in Majembe (Lulindi Division, Masasi District), died April 1992.
Ms Mariamu Rashidi Nankandila Nachinguru, born 1962 in Mnyambe (Kitangari, Division Newala District). Ms Zakia Selemani Namalenga Ntanda, born march 1957 in
Mnyambe (Kitangari Division, Newala District). Mr. Vincent Amandus Nnali Ntanda, born 1957 in Mnyambe
(Kitangari, Division Newala District). and:
Mr. Verentino Mohamed Nnisindwa, born 1966 in Nkudumba (Kitangari Division, Newala District).
Chindonde: Ms Mariamu Mshamu, born 1963 in Kwanyama (Namikupa Division, Newala District).
Chimaraba: Mr. Ali, born 1972 in Mgao (Mpapura Division, Mtwara District). Mr. Abdallah Ntondo, Msanga Mkuu (Mayanga Division, Mtwara
District).
Ms Mwanankulu Selemani, born 1941 in Mnolela (Lindi Region). Chimatambwe: Mr. Leonardi Petro Milanzi, born 1944 in Mgwagule (Mchauru
Division, Masasi District), died August 2003, parents were born in Lichehe which is about six kilometres from Mgwagule. Shimakonde: Mr. Laurent Lijama, born 1949 in Mihula (Mueda District,
Mozambique).
Mr. Moses, born 1971 in Mtwara, Tanzania (both parents born in Mozambique).
via Mr. Benjamin Leach (SIL), Mueda, Mozambique:
Mr. Cosme Victor Wakudimba, born 1965(?) in Nanenda (Mueda District, Mozambique).
Mr. Ernesto Ali Vitandeka, born 1965 (?) in Nanenda (Mueda District, Mozambique).
Shimakonde took place in Mtwara with two informants; two other informants gave their information in writing via Benjamin Leach.
The fieldwork was conducted in Tanzania from July 1991 until April 1992, in the Netherlands (Chindonde) in 11 sessions from September until December 1992, and again in Tanzania from September 1993 until March 1994.
$ERXWWKLVJUDPPDU EDFNJURXQGSUHYLRXVVWXGLHVDQG
WKHRU\
My personal interest in Bantu languages of East Africa first led me to the Comoro archipelago in the Indian Ocean. The language of the island Ngazidja (or Grande Comore), called Shingazidja, was the topic of the thesis with which I completed my studies in African Linguistics, Leiden University, the Netherlands. Then I got the opportunity to write a dissertation, joining the Research School of Asian, African, and Amerindian Studies (CNWS; its predecessor was the Institute for Descriptive and Comparative Linguistics, IBVTW) of Leiden University, attached to the department of African Linguistics. There were three reasons why Makonde was chosen as the topic of my dissertation. First, I myself wanted to do research in northern Mozambique, but since the civil war was still going on at the time, southern Tanzania was chosen. Second, my promotor, prof. Thilo C. Schadeberg, felt that knowledge of Makonde was needed in order to understand more of the Swahili languages which are spoken along the coast in Tanzania and Mozambique, the main topic of his research. Third, the university of Dar es Salaam had initiated the project “ The Linguistic Atlas of Tanzania” , a project in which the department of African Linguistics of Leiden University participated. The goal of the project was to make a descriptions of all languages spoken in Tanzania, and my contribution to this project was to be a description of the Makonde language.
Previous studies on Makonde of Tanzania contained word lists, notes, introductions and outlines of the language, as well as some articles mainly about tonal aspects. We here list the previous studies (as far as we know) in chronological order (for the full references, see the Bibliography): Steere (1876) &ROOHFWLRQVIRUDKDQGERRNRIWKH 0DNRQGH ODQJXDJH; Lorenz (1914) (QWZXUI HLQHU .LPDNRQGH*UDPPDWLN; Large
(s.d.) (QJOLVK.LPDNRQGH; Johnson (1923) 1RWHV RQ .LPDNRQGH; Whiteley (1951) ,QWURGXFWLRQWRWKHVWXG\RI0DNRQGH (this work is said to be in the Newala District
Office, but I did not get the opportunity to see it when I was in Newala); Nurse (1979) 0DNRQGH; Yukawa (1989) $ WRQRORJLFDO VWXG\ RI 0DNRQGH YHUEV; Odden
(1990) 7RQH LQ WKH GLDOHFWV RI 0DNRQGH &KLPDUDED &KLPDKXWD; &FRPPDQG RU HGJHVLQ0DNRQGH.
in Tanzania, and to gather information about the dialectal situation in the area. Both goals were adapted during research. The first informants with whom I worked for the first few months were not very consistent, and when I found an excellent informant on the north-western dialect Chinnima, I decided to write a grammar of that variant. About the dialect chart, I wanted to give short descriptions of all dialects, but that turned out to be too ambitious a goal. I collected some information on three other dialects in Tanzania as well as on one dialect of Shimakonde in Mozambique which is given in Appendix C.3. There possibly is a fifth dialect, Chimachinga, spoken north of Lindi in Lindi Region; apart from a couple of days asking around (also in Lindi), I did not have enough time for elicitation. As this Chinnima grammar also contains information about other dialects of Makonde, the title of this work now is “ A Grammar of Makonde” followed by “ (Chinnima, Tanzania)” to indicate the focus of our research.
The fieldwork was conducted in a way many Bantuists work: begin with the nominal classes, which is the starting point for studying phonology and morphology, and gradually advance to the verbal system. Very early during fieldwork, texts were collected and analysed with informants, and words were gathered for the vocabulary. The main focus of this grammar is on phonology (especially tone) and morphology, less on syntax and semantics. This is the reason why we did not subcategorize verbs, for example, nor give a description of the exact differences in meaning and use of the locative prefixes and demonstratives. We have not thoroughly studied the temporal, aspectual and modal meanings of the inflectional categories of the verb, which we call “ tenses” ; our labels are but rough indications of some (hopefully salient) part of their meaning.
This work is not written within a particular theoretical framework, but we have greatly benefited from concepts and notions of various theories, especially generative phonology, autosegmental theory, lexical phonology and prosodic domain theory. The adopted concepts and notions are used for our main goal which is a descriptive one, and we wish to distinguish and describe the various rules and processes. We also indicate that these rules and processes are ordered. This final point, a derivational type of description, comes from a generative tradition. It differs from the constraint-based, non-derivational approach taken in Optimality theory (and its extension, Alignment Theory), but it remains our purpose to describe the language in a way that is useful to linguists from all traditions.
- a H tone on the first tone bearing unit of the stem (6+)
- a H tone on the second tone bearing unit of the stem (6+)
- a H tone on the final tone bearing unit of the stem (6)+)
- a H tone on both the first and the final tone bearing unit of the stem (66)+)
There are also stems without H tones, and we call this the QR+ tonal profile. This QR + tonal profile can either be combined with a prefix with a H tone or with a prefix
without a H tone, just like the 6)+ tonal profile; the other profiles can only be
combined with a prefix without a H tone. The combination of the tone of the prefix and the tonal profile of the stem is called a tone group. All words of the major categories fall into one of seven tone groups that can be distinguished, exemplified below by trisyllabic CVCVCV-stems preceded by a CV-prefix.
WRQH
JURXS 3UHIL[+ WRQDOVWHPSURILOH
$ 66)+ CV-CVCVCV H H % 6+ CV-CVCVCV H & 6)+ CV-CVCVCV H & + 6)+ H CV-CVCVCV H ' QR+ CV-CVCVCV ' + QR+ H CV-CVCVCV ( 6+ CV-CVCVCV H
extension ±LO ‘do something for someone’ and the Final ±D ‘Present, Past, Perfective
or Non-Past tense’ to form the stem ÛKLPELOD µdig for’, and when also an object
prefix (or: object concord, OC) like ±YD ‘them’ is added, the macrostem ÛYDKLPELOD
‘dig for them’ is formed.
OC root extension Final
YD KLPE LO D
| | verbal base | | verbal stem | | macrostem
Other verbal prefixes, including tense markers, are added to form verbal forms. Below, the Infinitive marker NX is added to form the Infinitive.
ÛNXYDKLPELODto dig for them
In case of nouns and pronominal forms, (pro)nominal prefixes are added to (pro)nominal stems. Next to morphological processes such as imbrication of the Perfective Final, all sorts of phonological processes apply in the lexicon including prenasalization, palatalization and vowel harmony, as well as prosodic processes. These deal with phonological phenomena concerned with tone, intonation and vowel length, such as H Tone Assignment (see 3.4). The Infinitive belongs to tone group A, i.e., the marker does not get a H tone and the tonal profile of the stem is S1/SF, and the form is ÛNXYiKtPELOi (the Infinitive is the only tense where the object prefix
also gets a H tone by assignment). H H H
| | |
ÛNXYDKLPELOD
Post-lexically, words are concatenated at the very first stage to build sentences. Sentences are also built in two ways: syntactically, i.e., with syntactic phrases, and prosodically, i.e., with phonological phrases. Prosodic phonology cooperates with syntax to produce the surface forms of phrases and sentences. This cooperation does not mean that syntactic phrases and phonological phrases always have a one-to-one relationship. The examples below show that some VP’ s and NP’ s fall together with one phonological phrase (conjoint) while others consist of more phonological phrases (disjoint); a phonological phrase is recognized by penultimate lengthening of its final word which is underlined in the examples, the end of a phonological phrase is marked by the sign ‘_’ accordingly.
WXWyQJyOiFKtPiNyRQGH_ we spoke Makonde
WXWRQJZHOHFKtPiNyRQGH_ we have spoken Makonde
W~QiWRQJRyOD_FKLPiNyRQGH_ we spoke Makonde
YDO~PpYiiQR | these men
YDO~PpYHpWX_ our men
YDO~~PH_YiDQML_ other men
YDO~~PH_YDN~O~XQJZD_ big men
Penultimate lengthening is the first post-lexical process. It is easy to observe since there is no contrastive vowel length in the language. The other processes include retraction of the final H tone, H tone shift, H tone bridge, H tone doubling and tone assimilation. These processes are ordered, and they apply to all words individually in the p-phrase as well as to the whole p-phrase (ch.3). There are also processes which specifically occur in whole p-phrases and not to words individually, such as special Noun-Specifier rules, a variant of H Tone Bridge and Regressive H Tone Dissimilation. These phrasal rules apply together at a certain stage of the derivation (ch.8).
Phonology, morphology and syntax are to a great extent interconnected in Makonde, as is usual in Bantu languages. There is a phonology-morphology interface since many phonological processes must be described as occurring in morphological environments, like vowel coalescence and glide formation which apply where syllables meet. There is also a morphology-syntax interface, like the addition of verbal extensions as the applicative, passive or causative which may change the syntactic valence of the verb. And there is a phonology-morphology and a phonology-syntax interface as prosodic phonology cooperates with morphology and syntax in words and phonological phrases. In fact, the phonological phrase is the meeting point for phonology, morphology, syntax and pragmatics. The choice between a conjoint form (which is in the same phonological phrase as the following object) and a disjoint form (which is in a different phonological phrase as the following object), for instance, is determined by pragmatics in terms of focus and informational units.
the surface. Also interesting is an automatic rule which applies whenever the proper environment is met (lexically and post-lexically): Meeussen’ s Rule. It is an adjacency constraint, part of the OCP (Obligatory Contour Principle) group of constraints, which prohibits two primary H tones from appearing next to each other. This constraint is sensitive to morae, and since other tone rules work on morae as well, they are the best candidates for being the tone bearing units in Chinnima. Tonal rules apply within different domains, called here tonal domains: the syllable (e.g. Tone Assimilation), the stem (e.g. H Tone Bridge), the macrostem (e.g. H Tone Shift), the word (e.g. H Tone Doubling) and the phonological phrase (e.g. Regressive H Tone Dissimilation, to be dealt with in ch.8). When the domain of application is the (macro)stem, tone rules (like H Tone Assignment) have to take into account the syllable division as well. Other prosodic processes like Penultimate lengthening, Penultimate shortening and Structure simplification work on the (penultimate) syllable.
In chapter 4 to 6, nouns, pronominal forms, invariables and verbs are described. Their structure is shown, and their tonal properties are outlined. In the chapter about nouns (ch.4), the famous noun class system is explained. Nouns, like pronominal forms, consist of a prefix and a stem. The noun class prefix shows the class to which the noun belongs; there are 16 classes in Chinnima. Most nouns belong to a pair of classes, called a gender, one of which indicates the singular form, while the other indicates the plural form. The noun plays an important role in the agreement system, i.e., words, like pronominal forms and verbs, that depend on the noun for agreement take a prefix that belongs to the same class as the noun.
YDO~Pp YiiQR YDQLWyQJRyOD FKLPiNyRQGH
cl.2-man cl.2-this cl.2-TM-speak Makonde language these men have spoken Makonde
Also secondary, or outer, nominal prefixes like the locative ones are described, and their tonal behaviour shows their phonological status and the order of attachment in the lexicon. In chapter 5, pronominal forms like demonstratives, possessives and other forms that take the same kind of prefix are dealt with, together with invariables which remain unchanged. In chapter 6, the structure of verb forms is described in detail, with special attention to verbal extensions, suffixes behaving like them, and their relative order.
role in the conjoint/disjoint distinction. At the end of this chapter, we also pay attention to complex and compound tenses.
In chapter 8, phrasal tonology is dealt with. The tonal processes of longer p-phrases, noun plus specifier and verb plus complement, are described. Specifiers can be divided in a similar way as tenses into conjoint, disjoint and conjoint-disjoint. As syntax and prosody are not structured in the same way, the results of this description show some interesting facts about the phonology-syntax interface.
